HER2-positive breast cancers are highly aggressive, frequently resistant to chemotherapy, and associated with a high incidence of mortality. Therefore, therapy by alternative means, such as targeted delivery of therapeutic genes, may prove much more effective on these types of cancers than standard methods of treatment. The gene delivery system described here is novel in its utilization of recombinant adenoviral components which have been modified to achieve breast cancer cell-specific binding and delivery of conjugated DNA, while retaining the high efficiency cellular uptake features of the viral capsid. This system would require only the minimal adenoviral proteins necessary for cell surface binding and internalization, and lacks all other viral proteins and genes. Thus, this system avoids the concerns associated with using viruses such as mutation and viral recombination. The Specific Aims are to test the hypotheses that: 1. Recombinant soluble penton and fiber capsid protein translocation requires interactions of key molecular motifs found in capsid proteins with components of the cytoskeletal and intracellular trafficking machinery. This study will uncover useful information about the determinants governing breast cancer cell entry, as little is known about viral trafficking pathways in these cells. Immunofluorescence and confocal microscopy will be used here to visualize trafficking of wild-type and mutant proteins. Traffic-enhancing mutations will be incorporated into vector design and tested in Aims2&3. 2. Engineered capsid proteins mediate non-viral gene delivery to HER2+ breast cancer cells and are less toxic than whole recombinant Ad in vitro. Comparisons of cytotoxicity and gene transfer will be made between non-viral complexes and Ad. We will assess the contribution individual capsid proteins make toward the cellular response to Ad infection. 3. Engineered capsid proteins are less toxic and immunogenic than whole recombinant Ad, and can mediate targeted non-viral gene delivery to HER2+ breast cancer cells in vivo. As HER2+ cells overexpress the heregulin receptor, we examine of the capacity for heregulin to direct non-viral gene delivery to this tissue. We will assess the contribution of individual capsid proteins to the immune response to Ad, compare immunogenicity to Ad, and test a method of immune evasion.